My wife said "this is too good to keep to yourself" — so I translated my Croatian side project into a product

#webdev #nextjs #startup #showdev

The problem

For years I had a recurring problem at work. When equipment gets
written off — laptops, phones, gear — it's often still in great
condition and more people want it than there are spots available.

I tried everything:

Drawing names from a hat — people would "accidentally" draw again when no one was looking
A shared spreadsheet — easy to manipulate, timestamps unreliable
First come, first served — whoever lobbied hardest won

None of it felt truly fair. So I built a solution — for myself.

It started in Croatian

The first version was entirely in Croatian. It was a personal tool,
just for my workplace. I never planned to share it publicly.

Then I showed it to my wife. She's a professional translator, so
language is her thing. She looked at it and said: "This is really
good. Other people could have the same problem — translate it and make it
available to everyone."

That was the push I needed. So I translated everything, polished
the UI, and decided to launch it properly.

The solution

dibsl.com — create a timed event, share
a link, and whoever clicks fastest wins a spot.

Server-side timestamps — no cheating
Anonymous codes — transparent results without exposing emails
Waitlist — if a winner steps down, next person gets notified
Bot detection — clicks under 150ms are rejected
Campaign system — organise multiple drops under one campaign

Tech stack

Frontend: Next.js 16, Tailwind CSS
Backend: Supabase (PostgreSQL + Realtime)
Scheduling: QStash (Upstash) for precise event timing
Emails: Resend
Hosting: Vercel

The interesting technical challenge

The hardest part was ensuring fairness. Reaction time is calculated
server-side: reakcija_ms = server_click_time - actual_start_time

This means network latency is part of the game — just like in real
drops. But server-side ranking means no one can cheat with
autoclickers or scripts.